This page will look better in a browser that supports web standards, but its content is accessible to any browser or Internet device.

3.2: The View from the Air

According to Andrew Beall, remote sensing was an important step in coordinating the Brown Marsh Project from the very outset. "Given its large regional scale," Beall notes, "Landsat imaging can help the aerial photographic effort proceed more efficiently by identifying potential 'hotspots' that bear closer scrutiny."

Leading the way in the aerial photography and subsequent mapping efforts has been Larry Handley, a senior geographer with USGS. Overseeing the effort from the 60,000-foot flights to the hand-drawn polygons that are eventually rendered on maps themselves, Handley describes the work his team has done on the Brown Marsh Project as "pushing the envelope, the most detailed analysis we've ever done at the 1 to 24,000 scale." More than simply taking photographs and creating maps from them, the process has involved coordinating affiliated Brown Marsh Project investigators in order to determine the ground characterizations they wished to see and then, after interpreting the aerial imagery, creating maps that accurately depict these characterizations.

An aerial photograph of West Terrebonne Bay, one of the dozens of photographs the Handley team took as the starting points in their mapping efforts. Even at 60,000 feet, afflicted marsh areas can be readily distinguished from their healthier counterparts.

At the interface between ecology and mapping, the search for agreed-upon ground cover characterization standards is called "signature key development." To this end, one of Handley's first steps after acquiring the photography was to gather together researchers from across the various disciplines so that he could garner their viewpoints and design a signature key around their research needs.

Beyond the six classifications Handley would eventually put forth, this signature key would, in effect, come to serve as the common nomenclature among researchers. To encourage this, Handley codified various research groups' pre-existing classification schemes-such as Beall's four-fold scheme above-so that they could be "cross-walked" with respect to his. In addition, he held what might be called "norming" sessions among the various project participants, so that researchers-be they on the ground, in the air, or looking at maps-could consistently distinguish among the six categories encompassing healthy to completely dead tracts of marsh.

Of course, there were still the maps themselves to produce. This effort is a complex process that involves interpreting the photography through stereo-optics so that textural, elevation, and contour details can emerge, details that provide contextual cues to the interpreters that one-dimensional analysis cannot. After these cues are identified by the interpreters, the photographs are then superimposed onto existing maps of the region by means of a light table where the signature polygons are then painstakingly hand drawn, with the smallest of polygons representing areas as little as 40 square feet. These maps were subsequently "ground-truthed" (i.e., spot-checked for accuracy by teams on the ground) before being distributed, along with the copies of the photography, to all researchers on the Brown Marsh Project.

Closer to the ground by some 59,850 feet were the aerial surveys conducted by USGS wetlands ecologist Thomas Michot. A pilot with nearly 30 years experience flying the Louisiana coastline in his long-running waterfowl surveys, Michot took to the air less than a week after Linscombe's initial discovery, videotaping the various degrees of browning he witnessed along the coast. Less than two weeks after this initial reconnaissance flight, Michot was back in the air in early June 2000 with a survey protocol in place.

Flying with Chris Wells, a USGS geographer with the Handley mapping team, Michot flew the coast from the middle of the state eastward in what would become the first of six transects, or flight lines, he would establish. Employing a pre-signature key four-fold classification scheme (green; mostly green with some brown; mostly brown with some green; and brown), the pair flew at 150 feet, Michot observing from the pilot's side window while Wells peered through the passenger's side. Every five seconds, each would characterize the nature of the marsh they were viewing 45° beneath them by speaking into their headset mikes. These assessments were recorded on to two separate laptops, each of which was running a global positioning satellite (GPS) program as well.

Click to enlarge (205 KB)

The end result would be a map that recreated the flight, complete with the flight path, circular markers on the map indicating each time the mike had been opened for a comment, and an audio file of the comment associated with each marker. Over the course of the six transects, Michot and Wells would average roughly 2,500 assessments each, assessments that were then transcribed and tallied. With the aid of these initial maps, Michot was then able to accurately retrace his route along each of the transects, assuring that he and Wells were surveying the same terrain in their follow-up flight the following month.

Click to enlarge (176 KB)

Over two years later, the pair is still making near monthly flights over the same set of transects. In the wake of brown marsh tracts rotting and disappearing altogether, they were forced to add a fifth category of "unvegetated mud flat" as they tracked the spread of the dieback and, more recently, the extent of the natural recovery process. All told, the dozens of flights Michot and Wells have flown have provided a valuable confirmation of Beall's Landsat images and Handley's maps, along with providing a much more accurate means of correlating the air and ground trends than either of the higher platforms permit.

Flying less frequent helicopter surveys only 20 feet or so above the salt marshes is LDWF's Greg Linscombe, the wildlife ecologist with the Louisiana Department of Wildlife and Fisheries who first observed the brown marsh dieback in his nutria survey of May 2000. Using both input from Michot's flights and his early four-fold classification scheme, Linscombe initially flew a series of north-south transects between Four League Bay to the east and the Mississippi River to the west in August 2000.

With the transects roughly two miles apart and the areas assessed at .5 mile intervals, the flights were a time-consuming but necessary process, especially given that one of the task's central purposes was to map those areas with the most severe vegetative damage. Using GPS-linked computer software similar to Michot's, Linscombe confirmed that the Terrebonne basin was the most severely impacted region in the coastal zone. More important, perhaps, was the role Linscombe's data played in helping researchers select sites for the ground-based data collection, monitoring, and remediation efforts that would soon be getting under way.